For more than a decade, cybersecurity has sold one product under a thousand names: visibility.
SIEM for events. EDR for endpoints. ASM for the attack surface. CNAPP for the cloud. Exposure management for everything else. Every category promised the same thing: if we could just see enough, we would finally secure enough.
The visibility industry just got a wake-up call. Practitioners are no longer buying it.
Over the last several months, I have spoken with hundreds of CISOs, security engineers, and IT leaders. Different industries, different stack sizes, different maturity levels. The sentiment is almost identical, and there is a name for it: Diagnostic Fatigue.
Teams are not drowning because they lack visibility. They are drowning because visibility, on its own, has become indistinguishable from noise. Every dashboard adds another list. Every list adds another ticket. Every ticket adds another week to the backlog. The work of finding problems has industrialized; the work of fixing them has not.
Look at the math. Attackers operationalize a new exploit in roughly 51 seconds. The average enterprise takes 27 days to fix a single exposure. That is not a visibility gap. That is an action gap, and no amount of better dashboards will close it.
And it just got worse. This month, Anthropic released Claude Mythos, an AI that surfaced more than 2,000 previously unknown software vulnerabilities in seven weeks of testing, with an 83% success rate in turning them into working exploits on the first try. Anthropic put it in defenders’ hands too, through Project Glasswing. That is the right instinct, and a clarifying moment for the industry: when vulnerability discovery runs at AI speed on both sides, every day a fix sits in a queue is a day the asymmetry compounds. More findings, faster, do not help a team that still takes 27 days to patch.
The industry’s response to the action gap has been quiet but telling: we have accepted “Assume Breach” as a strategy. Assume the attacker is already inside. Assume detection is the best you can do. Assume prevention at scale is no longer realistic.
That is not a strategy. That is a white flag dressed up in a framework.
Assume Breach was a useful mindset a decade ago, meant to keep defenders humble. Somewhere along the way, it became permission to stop trying to prevent. We outsourced our ambition and called it maturity.
It is time to flip the script.
The next era of security will not be defined by who sees the most. It will be defined by who fixes the most, fastest, without breaking the business. That requires tooling that produces outcomes, not findings: validated remediations, simulated before deployment, executed across the controls a team already owns.
We do not need another pane of glass. We need a workforce. Specifically, an AI Security Engineer that can do what humans cannot do at scale: discover an exposure, plan the fix, simulate the impact, and apply it safely, in minutes rather than weeks.
Visibility was the last decade’s product. Remediation is this decade’s mandate. The vendors who confuse the two will keep selling reports. The ones who understand the difference will give defenders something the industry has not delivered in years: an unfair advantage.
It is time to stop managing security and start eliminating threats.
Written by Barak Klinghofer, Co-Founder and CEO of Reclaim Security.
Barak Klinghofer
Co-Founder and CEO of Reclaim Security